THE PROTEINS WITH A SINISTER SIDE

Rogue proteins that mutate, unheard of 10 years ago, may cause cancer and mad cow disease. Charles Arthur reports

IMAGINE the sort of answers you might have got if, a year ago, you had asked a random sample of people what the word "prion" meant. Some sort of electronic component? A make of car for travelling salesmen? They would be unlikely to answer with any confidence. It sounds like one of those vaguely remote, yet technical words: things you perhaps ought to know but don't.

That's all changed since March, when the government announced that the most plausible explanation for 10 cases of a new variant of the brain disorder Creutzfeldt-Jakob Disease (CJD) was exposure to materials infected with BSE, or mad cow disease. The prion is the putative link. According to current theories, it is the infective agent that causes both diseases.

The effects have been dramatic enough: the destruction of the pounds 500m British beef export industry, and the possibility of a CJD epidemic, for example. But the idea is revolutionary on its own. Ten years ago the consensus was that all diseases were caused by bacteria, viruses, or "viroids" - biological entities which contain their own genetic material (DNA or RNA). But the prion is quite different: it is simply a protein, one of the building blocks of the body. The best explanation at present for the cause of diseases like BSE and CJD, as proposed by Professor Stanley Prusiner (see box overleaf) in the Eighties, is that a protein called PrP - found in the brain and in all the body's cells - spontaneously "flips" from its natural, stable shape to another. The new abnormal shape is also stable, but useless - or even detrimental - to the body.

According to the theory, the misshapen protein then causes normal copies of the protein in the brain to change shape too, causing the neurons in the brain to lose functionality until the terrible effects become visible - holes in the brain tissue and a descent into coma and death.

Prusiner has had to fight hard to get his ideas accepted; a number of scientists still deny that he is right. Yet the idea of proteins which change shape is not new. It happens all the time inside us. We'd be dead without it: haemoglobin, the protein which carries oxygen in the blood, has different shapes depending on whether it is carrying an oxygen molecule or not. Other proteins in cell membranes use similar changes as cellular switches: there is an "operational" shape while they are allowing ions to pass, but a different one when preventing ion transfer.

But while the battle over prions has been raging in the world of sheep and cows and humans, other scientists have been bending their minds to a related question. If Prusiner is right, could other diseases also be caused by rogue proteins - the spontaneous change in shape from one which is functional to one which is not?

Questions like this can only be answered now that we are in what might be called the "third age" of organic chemistry. The first age - dating back to medieval alchemists - comprised understanding the basic chemical constituents of complex molecules: how many carbons, how many hydrogens, and so on. It told you about ratios: how much of this (say, sodium) would react completely with that (say, water) and what it would produce.

In that age, chemicals were simple lines of atoms. The second age became established in 1865, when August Kekule was puzzling over the structure of benzene - known to have six carbon and six hydrogen atoms in each molecule. But how could that be? Carbon atoms attach to four other atoms; hydrogen atoms to one. As a line, it didn't add up. Puzzling over this, Kekule fell asleep on a bus, and dreamt of a snake swallowing its tail. He woke up and realised he had solved his problem: the carbon atoms were joined in a ring. Simple, and yet revolutionary.

The discovery of the double helix structure of DNA in 1953 was another huge step forward - into an age where shape, not just structure, matters. Nowadays, biologists regularly have to consider the shape of enormously complex molecules.

Take p53 - a protein produced in every body cell. There are 393 amino acids in p53, chained together in a precise sequence. Each amino acid is itself a chain of carbon atoms and other elements. The complexity means that as it is generated (by the p53 gene) piece by piece within the cell, the protein "folds" in three dimensions. This shape gives rise to its function - part of which is to fit with other complex molecules, and especially the most complex we know, cell DNA. If a gene mutates, and puts a different amino acid at some point, that can change the final shape of the finished protein, and dramatically affect its function. Or the mutation can have no effect. Presently, it's unpredictable.

p53 isn't a random choice. It has an intriguing link to prions when searching for experimental evidence that an abnormally folded protein can disrupt the folding of its normal counterpart.

Professor Jo Milner, of the department of biology at the University of York, studies p53. In 1991, while at the University of Cambridge, she showed that a mutant form of p53 not only disrupts the folding of normal p53, but also causes the normal p53 to adopt the "mutant" protein's shape. Here, however, the similarity between p53 and prions stops. Milner also showed that mutant p53's "dominance" of the normal protein only happens when the two proteins are synthesised side by side. In the prion hypothesis, it is the pre-formed prion protein that is thought to convert PrP to the abnormal form. "As far as I am aware," says Milner, "prions are the only example of infectious proteins." (And even those are still theory, she adds - though today an increasingly strong one.)

The p53 gene, present as a pair in every cell, carries the instructions to make the p53 protein. This has two tasks in its normal form. First, it can prevent cell division while in its common shape - known as the suppressor, or "S" conformation. Following cell growth stimulation, the protein briefly changes shape, to the promoter, or "P", conformation. (The mechanism that induces this shape change is not clear, though a chemical signal is suspected.) The block on division is lifted; the chromosomes duplicate; the cell divides.

It also detects any damage in the cell DNA's 300 million or so nucleotides. This occurs on average about 5,000 times a day in each cell, caused by chemical pollutants and radiation such as X-rays and ultraviolet light. When damage is detected, p53 begins processes to either repair that damage, or (if the damage is too severe) send the cell along one of two paths to oblivion: either never to replicate, or to kill itself.

You don't need p53 to live; laboratory animals have been bred which don't produce it. But they succumb to cancer before their natural lifespan. Milner remarks, "It shows that p53 isn't essential to the growth and survival of cells, but that it's important to protect against the development of cancerous changes." Others, including Mike Kastan at the Johns Hopkins University in Baltimore, and Geoff Wahl at the Salk Institute in California, have shown that p53 can block the division of cells containing damaged DNA. In this way, p53 plays a crucial role in maintaining genetic stability.

That would be fine, except that sometimes the damage to the cell DNA affects the p53 gene, which can affect the folding of the p53 it produces. It is then unable to fulfil its normal function: it can't arrest growth and division of the cell, nor repair any other damage to the DNA. This loss of p53 protection means that genetic damage goes unrepaired and is passed on to daughter cells at division, leading to genetic instability - and cancerous progression.

More than half of all human cancers are linked with mutations in the p53 genes. In some cancers - such as lung cancer - the linkage is much higher. A key element in the recent finding by US scientists that smoking directly causes lung cancer was the demonstration that a particular component of tobacco smoke causes exactly the p53 mutation found in lung cancer cells.

When working properly as a tumour suppressor, the p53 protein seems to "recognise" the presence of broken or damaged DNA, and binds to specific parts of it while calling for the cell's repair crews. A flaw in the p53 will mean this binding cannot occur: "It's like removing the earth pin from a three-pin plug," says Milner.

However, p53 has hidden abilities connected to its shape. This discovery, like many others in science, owes something to serendipity. A team at the Weizmann Institute of Science in Israel were comparing a number of different p53 mutants. Cells containing the mutants were grown in incubators at normal body temperature, 37C. But one incubator had a faulty thermostat: it stayed at a steady 32.5C. The cells in it stopped growing.

But when they were put into an incubator running at 37C, the cells recovered and started dividing. Moshe Oren, head of the research lab, realised that one of the mutant p53 proteins was sensitive to temperature: at 37C it was unable to suppress cell growth, but at 32.5C it functioned like the normal protein, and arrested cell growth.

Subsequently, Milner showed that the folding of the mutant protein was also sensitive to temperature, and that its shape appeared to be normal at the lower temperature.

This suggested an interesting line of research. What would mutant p53 do to normal p53? To test this, Milner used a cell-free system to synthesise mutant and normal p53 side by side. The result was that the normal p53 was driven by the mutant protein to fold into the abnormal, mutant shape. Somehow, the mutant p53 was affecting the normal protein. Yet mixing samples of normal and mutant p53 protein in a test tube after they had been synthesised showed no shape-changing effect - a key difference from prion theory. (Such post-synthesis shape-changing has been shown with sheep PrP, for example.)

Mutant p53 is frequently more stable than its normal counterpart - surviving for hours, rather than the 10 or so minutes that the normal form does unless it is bound to DNA. This results in its accumulation in affected cells: high levels of p53 in tumour cells are a good indication that a mutation has occurred. These raised levels, and the long survival of mutant p53, may explain the fact that some cancer patients have an immune response to p53. But together with the high incidence of p53 mutation in human cancer, this has given some more imaginative scientists a fresh set of approaches to anti-cancer therapies, based on the immune system.

A number of laboratories are devising experiments to stimulate the immune system to produce antibodies which might selectively attack tumour cells that express mutant p53. Another approach is to introduce normal p53 back into cancerous cells using gene therapy - the replacement of the genes inside the cell with ones introduced from outside.

But perhaps the most novel and exciting approach relies on a virus to selectively kill cancer cells which have mutant p53. This work, developed by Frank McCormick and published in the 18 October issue of Science, starts with a common respiratory virus. Normally it infects cells and, in order to replicate within them, uses one of its own viral proteins (called E1B) to inactivate p53. Without E1B, the virus has little or no effect on normal cells. However, in cancer cells with mutant p53 the defective virus can replicate and kill the cell. As viral particles are released from the dying cancer cell they may infect surrounding tumour cells, replicate and kill again. But because the defective virus cannot replicate in cells with intact p53, its spread should be restricted to the tumour. It's an elegant solution, which just might work. Milner certainly hopes so. The idea of using p53 itself, rather than other treatments, is one that especially appeals to her.

"p53 is, after all, a potent, naturally occurring tumour suppressor," she says. "We're exploiting the body's defences against tumours. If you can get the immune system to target cells with mutant p53, that's the way to go - rather than killing half the healthy cells in the treatment too."

It's an approach very different from most commercial biotechnology, she notes. "That just repeats things going down a road that's already known. But if you don't know what the end result is, you're standing at a crossroads wondering where to go. I think sometimes people are too narrow-minded."

It's an interesting time, this third age, with its juggling of structure, and form, and shape, and the constant effort to comprehend something which has had millions of years to get its function refined as close to perfection as evolution can manage. When Milner talks about p53, she gets excited in the way that scientists occasionally do: there's that tone of wonder at how it all works so well and so economically. Well, can you name anything else with just 393 constituent parts that can recognise and stop cancer?

"It's a beautiful protein," she says. There's absolutely no doubting she means it. And the way it fits together is so thrilling. "It's all so common-sense. And so good." !

PRION PIONEER

THE SIMILARITIES between Jo Milner's work and that of Stanley Prusiner - who first formalised the "prion" theory for diseases like CJD and scrapie - are inescapable. It was pure accident, but Milner found Prusiner's work illuminating, and was delighted to meet him in 1991. She invited him to Cambridge when he made a flying visit to the country that year, just after a paper of his had been published further promoting his theory, which by then was beginning to win other converts.

"We sat around a table and talked for hours. My impression was of somebody fired with enthusiasm, highly articulate. He was teaching me for most of it. He was aware of all the arguments and had designed his experiments to be aware of that."

It's an interesting insight into Prusiner, who has a reputation among journalists who cover science as remote and uncontactable. This is partly his own choice: he very rarely speaks to reporters. According to a July report in Science magazine, Prusiner was particularly wounded in 1986 by an article headlined "The name of the game is fame. But is it science?" which appeared in the popular science magazine Discover. Since then, he hasn't opened up much - though he has continued working hard backing up at the scientific ideas he promulgates with clear experiments.

He has made the occasional appearance above the media parapet: he wrote an article for Scientific American in January 1995 detailing how he came to the concept of prions, and how this concept explained the particular characteristics of scrapie and Creutzfeldt-Jakob Disease (CJD), and especially inherited forms of it caused by genetic abnormalities.

Prusiner makes an interesting exception. Most scientists challenging the scientific orthodoxy use the media to push their message as hard as possible. Prusiner has chosen the other route: getting the science done.

It may be that in the long run this will actually be of benefit to him, since it has meant that the prion theory has had to prove itself in the laboratory.

So far, it has stood up to those tests extremely well.